Systems Engineering and Electronics ›› 2018, Vol. 40 ›› Issue (1): 225-230.doi: 10.3969/j.issn.1001-506X.2018.01.32

Previous Articles     Next Articles

Method of missing data imputation for multivariate time series

LI Zhengxin, ZHANG Fengming, WANG Ying, TAO Qian, LI Chao   

  1. Equipment Management and Unmanned Aerial Vehicle Engineering College, Air Force Engineering University, Xi’an 710051, China
  • Online:2018-01-08 Published:2018-01-08

PDF (PC)

Abstract:

Multivariate time series is a common data type. However, due to many kinds of interference factors, missing data are inevitable, which affects data process and analysis. Firstly, for the series containing missing data, similar series that are of the same class with it are searched to form the training set. Secondly, making use of least squares support vector machine, missing data are filled by univariate and multivariate filling methods respectively. Then, according to the difference between the filling results of univariate and multivariate filling methods, a combined method is proposed. Finally, extensive experiments are conducted. The results show that the proposed method can fill missing data precisely, and can be used in the case where the amount of missing data is relatively large.

[an error occurred while processing this directive]